gym-examples 3.0.754__py3-none-any.whl → 3.0.755__py3-none-any.whl

gym_examples/__init__.py

@@ -5,4 +5,4 @@ register(
      entry_point="gym_examples.envs:WSNRoutingEnv",
 )
 
-__version__ = "3.0.754"
+__version__ = "3.0.755"

gym_examples/envs/wsn_env.py

@@ -30,7 +30,7 @@ base_back_up_dir = "results/data/"
 max_reward = 1 # maximum reward value when the sensors sent data to the base station. The opposite value is when the sensors perform an unauthorized action
 
 class ScalarAttentionModel(nn.Module):
-    def __init__(self, input_dim=4, output_dim=1):
+    def __init__(self, input_dim, output_dim=1):
         super(ScalarAttentionModel, self).__init__()
         # Initialize GaussianAdaptiveAttention
         self.ga_attention = GaussianAdaptiveAttention(
@@ -51,8 +51,8 @@ class ScalarAttentionModel(nn.Module):
         scalar_output = self.output_layer(attention_output)
         return scalar_output
 
-net = ScalarAttentionModel(input_dim)
-net = net.double()  # Convert the weights to Double
+# net = ScalarAttentionModel(input_dim)
+# net = net.double()  # Convert the weights to Double
 
 class WSNRoutingEnv(gym.Env):
 
@@ -227,7 +227,7 @@ class WSNRoutingEnv(gym.Env):
 
         # rewards = [reward.item() if isinstance(reward, torch.Tensor) else reward for reward in rewards] # Convert the reward to a float
         
-        rewards = self.compute_attention_rewards(rewards)
+        rewards = self.compute_attention_rewards(rewards, self.n_sensors)
         print(f"Rewards: {rewards}")
         # rewards = np.mean(rewards)
 
@@ -466,16 +466,18 @@ class WSNRoutingEnv(gym.Env):
         return np.clip(normalized_throughput, 0, 1)
     
 
-    def compute_attention_rewards(self, rewards):
+    def compute_attention_rewards(self, rewards, n_sensors):
         '''
         Compute the attention-based rewards
         '''
+        net = ScalarAttentionModel(n_sensors)
+        net = net.double()  # Convert the weights to Double
         final_reward = []
         for i in range(len(rewards[0])):
             rewards_i = [reward[i] for reward in rewards]
             rewards_i = torch.tensor(rewards_i, dtype=torch.double)
             rewards_i = rewards_i.unsqueeze(0) # Add batch dimension
-            final_reward.append(net(rewards_i))
+            final_reward.append(net(rewards_i).item())
         
         return final_reward.mean().item() 
 

{gym_examples-3.0.754.dist-info → gym_examples-3.0.755.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: gym-examples
-Version: 3.0.754
+Version: 3.0.755
 Summary: A custom environment for multi-agent reinforcement learning focused on WSN routing.
 Home-page: https://github.com/gedji/CODES.git
 Author: Georges Djimefo

gym_examples-3.0.755.dist-info/RECORD

@@ -0,0 +1,7 @@
+gym_examples/__init__.py,sha256=815cyqeKkNC7ftFt8Vtj07athkj36c2STNlFxmRcqlc,166
+gym_examples/envs/__init__.py,sha256=lgMe4pyOuUTgTBUddM0iwMlETsYTwFShny6ifm8PGM8,53
+gym_examples/envs/wsn_env.py,sha256=h56qfOsV7mjaQi9eaZ1Mc92X322sGXPZV_KYUeUgiRE,27000
+gym_examples-3.0.755.dist-info/METADATA,sha256=2wawIwJy9MsQXABHUT0Oq63fUBw2mtHxKRw2Ipv-R_Q,412
+gym_examples-3.0.755.dist-info/WHEEL,sha256=2wepM1nk4DS4eFpYrW1TTqPcoGNfHhhO_i5m4cOimbo,92
+gym_examples-3.0.755.dist-info/top_level.txt,sha256=rJRksoAF32M6lTLBEwYzRdo4PgtejceaNnnZ3HeY_Rk,13
+gym_examples-3.0.755.dist-info/RECORD,,

gym_examples-3.0.754.dist-info/RECORD

@@ -1,7 +0,0 @@
-gym_examples/__init__.py,sha256=AfgfcIhRLQ13gj-B6lbC4iTEtYaTWCO14JRFDumF1dA,166
-gym_examples/envs/__init__.py,sha256=lgMe4pyOuUTgTBUddM0iwMlETsYTwFShny6ifm8PGM8,53
-gym_examples/envs/wsn_env.py,sha256=I2px13TJyNn93cCQkO22hI5WuIJo_HT3v0ZfBqk7_0k,26856
-gym_examples-3.0.754.dist-info/METADATA,sha256=OrruAdPUHtFXCbUzZNRWkDB8MHpR2zMNqpGNY5ZFHOw,412
-gym_examples-3.0.754.dist-info/WHEEL,sha256=2wepM1nk4DS4eFpYrW1TTqPcoGNfHhhO_i5m4cOimbo,92
-gym_examples-3.0.754.dist-info/top_level.txt,sha256=rJRksoAF32M6lTLBEwYzRdo4PgtejceaNnnZ3HeY_Rk,13
-gym_examples-3.0.754.dist-info/RECORD,,